GEOLOGICAL AND PETROPHYSICAL EVALUATION
OF THE PIONEER ANTICLINE AREA
T11N/R22W SBBM

PREPARED BY
DIGITAL PETROPHYSICS, INC.

William R. Berry II
Geologist/Petrophysicist

Deborah M. Olson
Geologist/Petrophysicist

David J. Crane
Geologist/Petrophysicist

UNDER SUBCONTRACT TO
MICHIGAN TECHNOLOGICAL UNIVERSITY

DOE CONTRACT NO. DE-AC22-94BC14892


TABLE OF CONTENTS

  1. SUMMARY
  2. INTRODUCTION
  3. REGIONAL GEOLOGY OVERVIEW
    1. TECTONICS
    2. STRATIGRAPHY
      1. NOTE ON DIATOMACEOUS LITHOLOGIES
    3. GEOLOGY OF THE STUDY AREA
      1. EXISTING FIELDS AND PRODUCTION IN THE STUDY AREA
        1. PIONEER FIELD
        2. LOS LOBOS FIELD
        3. OTHER DRILLING
      2. THE ETCHEGOIN AND REEF RIDGE/MONTEREY IN THE PIONEER AREA
  4. METHODS USED IN THIS PROJECT
    1. CYMRIC ANALOG
    2. LITHOLOGIC AND CORE DATA AVAILABLE
    3. ANALYSIS APPROACH FOR THIS STUDY
      1. OVERVIEW OF DPI PROJECT TASKS
      2. PETROPHYSICAL MODEL: DUAL WATER
      3. WORK SEQUENCE
      4. DATABASE PREPARATION AND EDITING
        1. DATABASE CONSTRUCTION
        2. ENVIRONMENTAL CORRECTIONS
        3. NORMALIZATION
      5. PARAMETER SELECTION
        1. CALCULATION OF CLAY VOLUME
        2. CALCULATION OF POROSITY
        3. CALCULATION OF WATER SATURATION
  5. RESULTS OF THE STUDY
    1. OBVSERVATIONS AND CONCLUSIONS
    2. OUTPUT DATA INCLUDED WITH THE REPORT
      1. FINAL REPORT
      2. FINAL DATABASE
  6. DISCLAIMER

APPENDICES

  1. ANALYSIS APPROACH
    1. PETROPHYSICAL MODEL
      1. CHOICE OF MODEL
      2. POROSITY MODEL
      3. SATURATION MODEL
  2. MONTEREY SHALE FACIES AND PRODUCTION CHARACTERISTICS
    1. FORMATION OF A DIATOMACEOUS RESERVOIR
    2. ANALOGOUS SILICEOUS SHALE RESERVOIRS IN THE SOUTHWESTERN SAN JOAQUIN BASIN
      1. REEF RIDGE FORMATION
      2. MCCLURE/ANTELOPE SHALE FORMATION
        1. CAHN ZONE OF THE MCCLURE/ANTELOPE SHALE FORMATION AT LOST HILLS
        2. MCDONALD SHALE RESERVOIRS AT LOST HILLS
    3. FRACTURING
      1. MICRO FRACTURING
      2. MACRO FRACTURING
    4. PRODUCTIVE POTENTIAL OF BIOGENIC SILICEOUS SHALE RESERVOIRS
      1. FRACTURE STIMULATED PRODUCTION AT LOST HILLS
      2. PRODUCTION ANALOGS IN OTHER FIELDS
    5. HIGH WATER CUTS COMMON IN BIOGENIC SILICEOUS SHALE RESERVOIRS
  3. DATABASE PREPARATION AND EDITING
    1. DATA LOADING
    2. DEPTH SHIFTING
    3. ENVIRONMENTAL CORRECTIONS
    4. NORMALIZATION
      1. NORMALIZATION DEFINED
      2. GR METHODOLOGY
      3. POROSITY NORMALIZATION
        1. DENSITY METHODOLOGY
        2. NEUTRON METHODOLOGY
      4. INDUCTION METHODOLOGY
  4. PARAMETER SELECTION
    1. CALCULATION OF VCLAY
    2. CALCULATION OF POROSITY
    3. CALCULATION OF WATER SATURATION
  5. FIGURES

1.0 SUMMARY

This study was performed under subcontract to Michigan Technological University as part of a grant from the Department of Energy to demonstrate computer analysis and visualization techniques for a reservoir description project. The Pioneer Field area, located approximately 10 miles southeast of the town of Maricopa in the San Joaquin Valley, California, was chosen for the example study. This area is located on the plunging nose of an anticline and therefore has considerable structural relief. The project focused on the upper Miocene siliceous shales, but the overlying shaly sands of the Etchegoin Formation were also included in the geological and petrophysical analysis.

The study area was good for demonstration purposes for a variety of reasons. The shaly sand and siliceous reservoirs are typical of the more challenging lithologies present in California, and that are most difficult to characterize and visualize for operators now facing redevelopment or EOR activities to maintain production. The combination of new wells and old E-logs present at Pioneer is quite typical of California's oil fields, and part of the results of this study is a technique for using these data as part of a reservoir evaluation. The area is located on an anticline with high relief and major faults, which is often difficult to visualize from 2-D paper cross sections and maps, so 3-D visualization techniques are quite applicable here.

The diatomaceous reservoirs of the Monterey Formation consist predominantly of siliceous, cherty shales containing various amounts of clay and silt. These reservoirs have high porosities and low permeabilities. The key to production in the Opal-Ct and quartz phases is brittleness and fracture permeability. Most of the siliceous section at Pioneer is in the Opal-Ct to quartz phase, which is at a higher diagenetic grade than its current depth would allow. This indicates that the Miocene was at one time buried deeper than it is now.

The petrophysical and geological analysis that has been performed as part of this study, and that is discussed in this report, resulted in a reservoir description including lithology and porosity for all wells, and water saturation for those wells with adequate logs. Additionally, a methodology is presented for using the wells with porosity logs to calibrate an empirical model for deriving lithology and porosity from the E-log only wells.

2.0 INTRODUCTION

The main objective of the project is to provide the smaller oil field operator with the approach and technology required to undertake an E.O.R. reservoir characterization and evaluation similar to those performed by larger international companies.

The scope of work outlined in the project proposal is categorized into several interactive and parallel phases.

. Database Initialization and Management

. Log Calibration and Analyses

. Rock Measurement

. Petrophysical and Geological Modelling

The database management phase provided the vehicle for integrating all available data and interpreted results into the project to enable the process of visualization. This phase involved numerous related tasks including data gathering, digitizing, and construction of an interactive multi-media database. This phase of the project was ongoing and interactive throughout the entire study.

The log calibration and analyses phase involved a detailed review of all of the available log and other well data, and capturing these data digitally for use in the various parts of the project.

The rock measurement work involved the use of various standard and experimental analysis techniques on core and cuttings samples from the project study area to better determine rock properties. The results of the experimental techniques were reviewed and compared to discover whether these methods would have further application in California and elsewhere to measure rock properties.

The petrophysical and geological modelling work included the log analysis, correlation, and mapping work for reservoir description. Preliminary correlations and maps were drawn using unprocessed log data, and were refined later using the results from the log analysis. Physical and chemical properties gathered from laboratory measurements of conventional core and cuttings were also used to develop algorithims which related geological and engineering parameters to log responses. Some core analysis had been performed previous to the commencement of this project, and other tests were done as part of this work. Well log, core and other test data were assembled and stored in QuattroPro and Excel, commercial spreadsheet packages, which allowed for easy access and updating and also for uploading to the project database. Log editing, preparation, and plotting were done at DPI in the Petrolog system from Crocker Data Processing, and the analysis was done in Symbiolog. Cross sections and maps were drawn at MTU using GeoGraphix.

The tasks completed by DPI included:

Acquisition of all log, core, and well history data available within the study area Digitizing and data capture of well data Construction of log database Analysis of field wells Correlations of formations and markers

This report discusses primarily the work done by DPI on this project. Also included is general information about the geology of the study area and properties of the diatomaceous rocks that are the principal subject of the work.

3.0 REGIONAL GEOLOGY OVERVIEW

The San Joaquin Basin is roughly coincident with the southern portion of the geographic San Joaquin Valley (or Central Valley) of California. It is a downwarped basin with a total thickness of over 35,000 ft of primarily marine sediments ranging in age from late Cretaceous to Recent. The basin is bounded to the east by the Sierra Nevada Mountains, and on the south and west by the San Andreas Fault and associated structures. The northern boundary of the basin is a subsurface rise called the Stockton Arch. The east flank is characterized by gentle dips and primarily normal faulting. The south and west flanks are characterized by steep dips and compressional tectonics.

3.1 TECTONICS

Plate margin tectonics have always dominated the geologic development of western California. Although evidence in the form of remnant metasedimentary bodies exists for Paleozoic continental-platform type deposition in California, the earliest recognizable coherent rock units belong to a subduction-forearc basin compex that formed during Mesozoic time. Most of the "basement" rock west of the San Joaquin Valley consists of a melange of metamorphosed subduction-related rocks called the Franciscan Formation. The basement to the east of the central valley is granitic, comprising the roots of the Mesozoic island arc that bounded the subduction zone. It is not certain where the boundary, and therefore the continental margin, is located beneath the sedimentary cover in the San Joaquin Valley. The oldest recognizable sediments are sand-rich Cretaceous turbidites deposited in the forearc basin. Turbidites and related deep-water deposits dominated the sedimentary section until Pliocene time.

In early Miocene time, the oceanic spreading center that was moving eastward with continuing collided with the subduction zone and became a transform fault. This fault or system of faults, collectively called the San Andreas, grew progressively longer as the triple junctions moved northward and southward along the coast. This initiated a regime of oblique compression along the western margin of the basin. This active tectonic system has been the most significant influence on the structural and stratigraphic development of the basin from mid-Miocene time to the present.

Several sets of en-echelon anticlines have formed along the west and south margins of the basin, all of which contain oil fields. Because of the rapid syn-depositional structural growth, the sedimentary sections are characterized by rapid facies changes and many local and regional unconformities.

3.2 STRATIGRAPHY

The stratigraphy of the San Joaquin Basin is shown in Fig. 1. This table was taken from Maher, Carter and Lantz (1975), and the reader is directed to this publication for an excellent summary of the geology and stratigraphy of the southern San Joaquin Valley. Formation names used in this table reflect formal U.S.G.S. usage; as would be expected, there are many informal local names used for these units throughout the basin.

The sedimentary section is dominated by the presence of turbidite sands and associated deep-marine shales. Most of the late Miocene-aged shales are diatomaceous, as are parts of the Eocene shales. These rocks are the source for most of California's oil, and under the right conditions also form reservoir rock. Late Miocene diatomaceous (or siliceous) shales are the the subject of this study in the Pioneer Anticline area.

3.2.1 NOTE ON DIATOMACEOUS LITHOLOGIES

Diatoms are part of the marine plankton that make their shells from hydrated amorphous silica (opal). Under certain conditions, their remains can dominate the pelagic sediments. These conditions prevailed throughout most of California's sedimentary basins during most of the Tertiary Period.

In deep basins that are anoxic at depth, the diatoms accumulate in fine millimeter-scale or thinner layers, and the organic material is preserved. With burial and heat, the amorphous opal (opal-A) of the original diatoms is converted progressively to opal-Ct (opal-cristobalite) and eventually to quartz. There is a commensurate decrease in rock volume and porosity and an increase in brittleness. Original opal-A has a real rock porosity of 60% or more, since there is porosity inside the diatom shells as well as around them. Opal-Ct has porosity ranging from 35%-45%, and quartz phase rocks can have porosity from 0% to about 25%. None of these phases has much primary permeability, although fractures in brittle rock provide abundant secondary permeability. With these ranges of porosity, diatomaceous rocks can have tremendous storage potential, and by using modern methods of well completion and stimulation Monterey-type rocks all over California are good producers. Diatomaceous rocks are discussed in more detail in Appendiz II.

3.3 GEOLOGY OF THE STUDY AREA

The Pioneer anticline is located five miles to the southeast of the town of Maricopa on the west side of the San Joaquin Valley. The east-west trending anticline has surface expression through areas of Sections 26, 27, 28 and 29 in T11/R23W. Both flanks are steeply dipping and the top is relatively flat, with an east-southeast plunge. Several thrust faults trending sub-parallel to the structural axis have been mapped on the north flank, including one that cuts several wells and causes a repeated section. Additionally, several cross faults were recognized that probably have an oblique-slip sense of motion.

The local names in use in the study area are noted on the Table 1 and shown in the stratigraphic column for the Pioneer Field taken from the DOG (Fig. 1a). Stratigraphically, the deepest wells in the study area are within the Pioneer Field and penetrate several thousand feet of the Temblor Formation. The overlying Monterey, Reef Ridge, and Etchegoin Formations were the study intervals for this project. The San Joaquin and Tulare Formations are present, but were not studied.

The Temblor Formation on this structure consists of a thick (at least several thousand feet) accumulation of sands and interbedded shales with a bedding thickness of ten to fifty feet. The trap for the Pioneer Field and other Temblor accumulations in the area consists of cross faults perpendicular to the anticline that juxtapose interbedded shales against sands, creating a series of thin stacked oil columns distributed over a large interval. Overlying the Temblor sands are a series of clay shales and silty sands that may represent the Gould and Devilwater Shale equivalents. Above this is a section that is about 1500 feet thick in the Pioneer Field area itself that consists of several intervals of siliceous shales interbedded with clay shales representative of the Monterey Formation. Down structure this section thickens and probably picks up some overlying Reef Ridge Formation, although it is difficult to tell the Monterey and Reef Ridge apart without paleontological assistance. The post-Miocene unconformity cuts the Monterey/Reef Ridge section, and it is probably deeply eroded near the structural crest. Overlying the unconformity are silty sands and gray clay shales of the Etchegoin Formation. The Etchegoin greatly thickens downdip. Another unconformity tops the Etchegoin, and this is overlain by the undifferentiated coarse sands and interbedded shales of the San Joaquin and Tulare Formations.

Structural growth has had a profound influence on the stratigraphic section. There is an angular unconformity present at the top of the Reef Ridge Formation, part of the widespread post-Miocene unconformity. Another unconformity is usually mapped at the top of the Etchegoin Formation. Substantial section must be missing across one or both of these discontinuities, since the diatomaceous rocks of the Monterey and Reef Ridge Formations are at a higher diagenetic grade than their present depth of burial would normally allow. This is discussed further in Section 4.2 and Appendix II. As a result of structural deformation, the oil-bearing siliceous shales of the Monterey and Reef Ridge Formations subcrop beneath younger rocks on the anticline. The oil that is produced throughout this part of the basin from Pliocene rocks is believed to be sourced from these subcropping Miocene shales. Pliocene outcropts to the west of the crest of the structure have been mapped as having oil stain adn tar on bedding planes. This oil very likely migrated from the aforementioned subcrops.

3.3.1 EXISTING FIELDS AND PRODUCTION IN THE STUDY AREA

3.3.1.1 PIONEER FIELD

The Pioneer Field itself includes about half a section of productive acreage in Section 32, T11N/R23E. It was discovered in July 1959 and as of the end of 1993 had produced a cumulative of 194,488 bbls. of 37-38° API oil from the Temblor Formation. The reservoir sands are tentatively assigned to the Media/Carneros equivalents. Two wells are still capable of production, of which one is currently on, averaging approximately 7 BOPD with a 16% water cut.

A small amount of Miocene production has been obtained in the Pioneer Field and vicinity, although most of the wells were not formally assigned to the field. In 1980 and 1981, Tenneco drilled six wells on the north flank of the anticline, about two miles northwest of the old Temblor producers. These wells were completed in the Monterey for a low daily rate of flowing, gas-charged heavy oil production. The production was uneconomic and the wells were idled. In 1990, another well was drilled in a different structural position, near the old Temblor wells on the crest of the structure. This well encountered similar conditions to the six Tenneco wells, and is still being produced at low rates.

3.3.1.2 LOS LOBOS FIELD

The Los Lobos Field, located about one mile north of the Pioneer Field, was discovered in 1952. Because of the 70o dips on this flank of the anticline, the wells are distributed in a line one well wide and about two and a half miles long. On the west end, it produces from a turbidite channel sand in the upper Reef Ridge Formation and on the east end from a sand in the overlying Etchegoin Formation. Productive depth ranges from about 6000 ft to nearly 8000 ft over a narrow horizontal range because of the dip. The field is now abandoned.

3.3.1.3 OTHER DRILLING

A number of exploration wells were drilled in the Pioneer area in the 1930's through 1950's, resulting in the discovery of the Pioneer and Los Lobos Fields. More drilling followed during the 1950's and 1960's, but no further discoveries were made. In the 1970's and 1980's, Tenneco (who held nearly all the land in the area) drilled several deep tests which were quite useful to this project. Many wells have had shows in the Monterey and in some cases were tested. A summary of shows in the more important wells is shown in Table 2.

3.3.2 THE ETCHEGOIN AND REEF RIDGE/MONTEREY IN THE PIONEER AREA

Although the main focus of the project was the siliceous shales of the Reef Ridge/Monterey section, the overlying Etchegoin Formation was included in the analysis. In this area, the upper part of the Etchegoin consists of a gray to gray-green clay shale overlying silty sands interbedded with shales. The section is 100 feet or less in thickness on the crest of the anticline, and thickens down the flanks. Changes in thickness result from a combination of erosional removal from the top and syndepositional growth of the anticline. Etchegoin sands are productive in the eastern part of the Los Lobos Field.

The Reef Ridge/Monterey section includes a channel sand in at the top that is present only in the western part of the Los Lobos Field, where is formed the main producing interval. Elsewhere, the lithology consists of fairlly clean diatomaceous rock interbedded with more clay-rich shales.

The major correlations of formation tops were readily visible on the unprocessed logs, and the top Etchegoin, top Miocene, Reef Ridge channel sand, and top Temblor were all identified prior to the availability of the petrophysical results. These tops were needed to assist in parameter selection and zonation of the wells for the log analysis phase of the work. Within the Monterey itself, however, few internal correlations could be recognized. In the past since the siliceous shales are difficult to correlate from raw paper logs, workers have picked the top Miocene unconformity upon which to base all their mapping, although it was well known that the pre-unconformity Miocene structure probably was quite different. If the shales could be correlated internally, structures could be identified that might lead to the discovery of new pools.

In the wells from the Pioneer area, it was possible to identify one correlation in the Miocene from the raw logs: the point labeled "R" on the type log (Fig. 1b). This is the top of an interval where the resistivity is more "spiky" than it is in overlying intervals. This probably indicates an increase in carbonate stringers, one of the typical Miocene lithologies. Above the "R" marker, however, it is difficult to trace correlations on raw logs. The additional correlations shown on the type log were made using the computed results. They show that the intervals of cleaner and shalier siliceous rocks can be traced throughout much of the study area, although their character changes somewhat from well to well. A stratigraphic cross section hung on the top Miocene shows these correlations in an approximate strike view along the north flank of the structure. Fig. 2 uses only raw logs, and Fig. 3 uses the computed logs.

The exercise of correlating the Miocene shows the value of using both the raw logs and the computed results for geological correlations. It is necessary to perform preliminary correlations on the raw logs, since formation/lithology boundaries are used in the petrophysical parameter selection, but it is important to go back and refine the correlations by using the results from the log analysis.

4.0 METHODS USED IN THIS PROJECT

At the start of this project, a data search revealed that no core material remained in existence for the Monterey section within the study area, although core analysis data from one well and cuttings from another were available. Therefore, to obtain rock material for experimental analysis, the project obtained permission from Unocal Corporation to utilize the core and logs from two wells within the diatomaceous reservoir in the Cymric Field, 25 miles to the northwest.

4.1 CYMRIC ANALOG

Wells in the Cymric field which were completed in the siliceous shales were reviewed in terms of completion technique, production performance, hydrocarbon concentration and reservoir rock quality to determine their suitability for use as an analog for diatomaceous production from the Pioneer Field area.

The Reef Ridge reservoir in the Cymric Wellport pool had produced a cumulative of 961,819 bbls. at 12.6° API oil and 13.38 mmcf gas as of 1993. Current production is 2,653 BOPD, with an average 72.6% water cut. Since discovery in 1957, 151 wells have been drilled to an average well depth of 3500 feet. Fifty-six wells are on production and 55 wells are shut in. This reservoir is somewhat unusual for Monterey-type production in the San Joaquin Valley in that it produces heavy oil from both the Opal-A and Opal-Ct phases. Most production elsewhere is from either the Opal-A or quartz phases, and is generally 20o API or higher. The Cymric Field is a good analog for the Pioneer Field production since the Monterey at Pioneer is also in the Opal-Ct phase and produces heavy oil.

The detailed core analysis and log data from the Cymric cores were used in setting up the analysis model for the Pioneer area. The most extensive Cymric data in Well #418 included a long section of Opal-A phase rock, some Opal-Ct, and the transition between them. The data from the deeper Well #415 included spot cores down to nearly 5000 ft. However, the lithology penetrated by the #415 was more clastic than diatomaceous in the lower intervals and therefore quartz-phase diatomite was not represented in the available core samples. The rocks at Pioneer are primarily in the Opal-Ct to quartz phases, with only a small amount of the section in the lower Opal-A phase range. However, the Cymric data were used to calibrate the Opal-A end of the model.

4.2 LITHOLOGIC AND CORE DATA AVAILABLE

Much of the lithologic data for the project came from the extensive analyses, descriptions, photographs, and personal inspections by team members of the Cymric analog core wells, Unocal #415 and #418. This included thin section and SEM photographs work done at MTU as part of the project. However, although no core material could be located for wells in the Pioneer area, reports on analyses done on the core from Tenneco #62X-30 were available. Additionally, Tenneco performed some detailed analyses on well samples, probably sidewalls, taken from two other wells in the area. Summarized below are the results from the Tenneco special analysis study.

Eight samples of the Antelope Shale taken from various wells in T11N/R23W were examined by Tenneco in 1980. The objective of the examination was to determine the mineralogy and degree of natural fracturing for each well.

The wells were located approximately 1.7 miles apart and were at structurally different depths. Results of the analysis indicated significant lithological differences between each well.

The samples from the Santiago (A) 72-30 well, in which the top Miocene is located at -257 feet subsea, consisted of a diatomaceous siltstone showing very little structure or fabric. Localized concentrations of detrital material included quartz, feldspar, altered rock fragments and detrital clay. Volcanic glass and calcareous microfossils were also identified, with pyrite fillings developed within some of the calcareous microfossils. Siliceous microfossils and glass shards are present in some samples and show no indications of dissolution.

The samples from the Santiago "C" 55-28 well, in which the top Miocene is at -2258 ft subsea, in contrast, was comprised of very finely laminated shale. Concentrations of pyritized diatoms other siliceous microfossils were absent or rare. Numerous microfactures dissected some of the samples but were healed by an isotropic fracture filling identified as amorphous silica. Some fracture fillings were composed of cristobalite. Open microfractures were not observed.

In summary, the samples from the Santiago "C" wells are very different from those of the Santiago "A" well. The "C" well is more strongly cemented which has resulted in the sealing of any microfractures which could have existed in the "C" well. Similar fractures are open in the "A" well. These naturally-occuring fractures are high-angle and indicative of compression. The fractures are closely-spaced and numerous and are the main conduit for hydrocarbon production. In contrast, the "C" well exhibits significantly less fracturing. High angle compression fractures are present, but more widely spaced. However, the rocks in the "C" well are at a more advanced stage of diagenesis and are more brittle, so continued tectonic movements are likely to continue fracturing the rock.

The samples were subjected to both acid exposure and simulated steam injection tests. Samples from the Santiago "C" well were very resistant to acid attack and remained intact, while the "A" well samples experienced various degrees of mechanical degradation.

The simulated steam injection tests had no effect upon the Santiago "C" samples and showed no signs of hydrocarbon displacement. Santiago "A" samples, however were broken down by steam injection with the maximum breakdown occurring in samples at depths between 1377 and 1394 feet. It was concluded that the degree of breakdown is directly related to the volume of smectite in the fine size fraction.

Potential reservoir problems were identified as a result of the laboratory tests and are summaried as follows:

1. The changes in mechanical strength of the Antelope shale is a direct result of variations in silica cementation. Santiago "A" 72-30 has low strength and will have low embedment pressure. Packing of induced fractures will present difficulties. The strength of the shale in the Santiago "C" 55-28 well will be high, as will the embedment pressure.

2. Acid sensitivity in the "A" well suggests potential for damage due to the secondary precipitation of iron compounds.

3. Fresh water sensitivity will be a major problem in Santiago "A" due to smectite clay. Steam injection tests demonstrate a large degree of disaggregation. The properties of the Antelope shale in the "C" well will be unaffected.

The lithology of the siliceous intervals in both wells is typical of the diatomite in this area, with clay content varying from 10% to 30%.

4.3 ANALYSIS APPROACH FOR THIS STUDY

DPI has established a procedure for performing petrophysical field studies that was used to organize and guide the work on this project. This procedure is discussed below.

A modern petrophysical field study that is properly performed has very significant implications for reservoir management. Although advances in tool design and the development of new analysis techniques have improved our ability to assess reservoir potential, the most important aspect of a modern field study is an integrated analytical approach.

The elements of an integrated analysis are:

- Development of a predictive model for the entire field (field function) - Preparation of a standard dataset for uniform processing - Model parameter selection based on reservoir geology - Engineering and production data incorporated where known - Use of a detailed plan for all studies that is adapted to specific project needs

DPI has used an integrated analytical approach for the Pioneer Anticline study. The elements are further discussed in Appendix I.

4.3.1 OVERVIEW OF DPI PROJECT TASKS

The primary purpose of this study was to perform a reservoir description to be used as an example in demonstrating the use of PC-based evaluation and visualization software. As part of this effort, DPI developed a petrophysical model that would be predictive of water saturation, porosity, shale, and clay volume in the Pioneer area. DPI also performed stratigraphic correlations and structural mapping to provide a basis for computerized visualizations of the structure and stratigraphy.

Well data were acquired and digitized. Since no core material was available from the Pioneer area, the project secured the release of extensive core material and analysis data from two wells in the Opal-A/Opal-Ct heavy oil reservoir in the Cymric Field. These cores provided samples for laboratory analyses at MTU and also calibration data for the Pioneer analysis model.

At DPI, the well data were edited, depth shifted where applicable, and standard environmental corrections were applied. Log normalization was done, then a foot-by-foot computerized analysis was run on all project wells that included a significant section of Etchegoin and Reef Ridge/Monterey Formations. For the wells that had modern resistivity and porosity logs (full log suites), it was possible to use standard petrophysical models to derive values of lithology, porosity, and saturation. Many project wells had only old electrical logs, however, and it is not possible to apply standard models to these wells. Based on the analytical results in the full suite wells, DPI developed an empirical model to derive lithology and porosity for these wells. It is not possible to compute a resistivity-based water saturation value from an electric log, since the old deep-reading curve is strongly asymmetrical in its response and standard machine processing is invalid. Therefore, no saturation calculation was performed for these wells.

These data were used to test various computer software packages for their usefulness in a small-company setting to perform reservoir description with PC-based tools. Most of the software testing was done by MTU staff using data generated by DPI and by MTU.

4.3.2 PETROPHYSICAL MODEL: DUAL WATER

Conventional interpretation models have not proven reliable for quantifying porosity, water saturation, and clay volumes in most San Joaquin Valley oil fields. Classical models such as Simandoux, Dispersed Clay, Laminated Clay, or Indonesian have yielded erroneous results in similar reservoirs due both to inadequacies of the models and improper parameter selection. The Dual Water model, while theoretically sound, can also yield erroneous results if parameters are improperly selected.

The model used in this study is unique and proprietary to DPI in that published equations have been applied in software written by DPI and in other systems that are commercially available. The model is based upon interpretation models generally accepted throughout the industry. The sequence of interpretation steps within the model is crucial to an accurate computation. Improper sequencing of otherwise correct equations will result in error where measurements require both shale and hydrocarbon corrections. Clay volume is first computed from each desired indicator. The minimum, median, or average clay volume is then determined as desired. Clay-corrected porosity is computed and iterated for hydrocarbon corrections. The hydrocarbon saturation of the invaded zone is computed on each iteration. Water saturation is then computed either by a quadratic form of the equation or by using an iterative method where 'n' is not equal to 2.0. The formation factor porosity exponent 'm' in the saturation equation is not a constant in the Pioneer area model, but is a variable as a function of lithology.

The model is discussed further in Appendix I.

4.3.3 WORK SEQUENCE

The work sequence followed in this project conforms to the Field Study Management template prepared by DPI. The only exceptions to the guidelines occurred when certain project-specific technical issues required further investigation outside the scope of the general Field Study outline. The main segments of the sequence are:

- Database preparation and editing - Environmental corrections - Normalization - Log correlation - Parameter selection - Verification of analysis and sensitivity runs - Identification of analysis problems - Final analysis - Verification of correlations with analytical results - Zone summations - Mapping of zone tops and properties from the zone summations - Final report

4.3.4 DATABASE PREPARATION AND EDITING

The database construction consists of: locating well data, creating tracking spreadsheets for each well, digitizing paper logs, editing data, depth shift and verification, environmental corrections, and normalization. The database was prepared for both the petrophysical analysis and for the geological interpretation.

4.3.4.1 DATABASE CONSTRUCTION

Pre-existing digital data were not available on any wells, therefore the paper logs were digitized. The curves were edited to remove invalid data, and then depth shifted where necessary. Header data and drilling zone information were also entered for each well.

Standardized curve nomenclature was used throughout the database preparation phase to ensure the maximum efficiency of batch processing later. A lexicon of curve names is included in Table 2. It is important, however, to retain the actual name of resistivity curves so that the type of tool can always be identified. A second set of resistivity curves was created in each well that consisted of copies of the original curves with uniform names for use with cross section programs.

4.3.4.2 ENVIRONMENTAL CORRECTIONS

The appropriate chartbook corrections were applied to the induction, gamma ray, neutron, and density logs. No similar corrections exist for old electric logs.


4.3.4.3 NORMALIZATION

Normalization was made difficult in this study by the large area, the structural and lithologic changes, and the scatter of the well data. A full discussion of normalization procedures is provided in Appendix III, Section AIII.1.4.

4.3.5 PARAMETER SELECTION

Parameter selection was challenging for this project due to the need to develop a useful model both for full-suite wells and for E-log only wells. Parameter selection proceeded in three stages: calculation of clay volume, calculation of porosity, and calculation of water saturation. The parameter selection process is discussed briefly in the following sections and in Appendix IV.

4.3.5.1 CALCULATION OF CLAY VOLUME

Volume of clay was computed by applying a non-linear transform to the baseline-shifted SP curve. For a few wells with poor SP's, clay was computed from the shallow resistivity curve. Very little quantitative data were available to calibrate the clay transform, so it was set up principally by the interpretation of descriptive data from core and mudlogs, local knowledge, and geological expertise. Details of this procedure are discussed in Appendix IV, Section AIV.1.1.

4.3.5.2 CALCULATION OF POROSITY

On full-suite wells, total and effective porosity were computed from the neutron-density crossplot after application of an iterative light-hydrocarbon correction. The model was calibrated to whole core porosity and verified with available sidewall core data. All core porosities are assumed to represent values close or equal to total porosity, since the samples were analyzed using the Dean-Stark procedure. No data were available to correct the measured laboratory data for the effects of overburden, so whole core porosities tend to be somewhat higher than computed values. For the E-log only wells, an empirical depth function was developed for total porosity based on the computed values from the full-suite wells. Effective porosity was then computed as a function of total porosity and clay volume.

Total porosity is computed from the crossplot by projecting the data points onto the matrix line along a certain slope, defined by a line between the matrix point and the dry clay point.

Details of the porosity model calibration are presented in Appendix IV, Section AIV1.2.

4.3.5.3 CALCULATION OF WATER SATURATION

After establishing good Vclay and porosity transforms, calculation of water saturation was performed. Some core analysis and mudlog data were available to assist in model calibration. Water saturation results are only available for full-suite wells, since the old lateral logs in E-log only wells are invalid for the calculation of a foot-by-foot water saturation. Methods exist for hand calculating the correction for the lateral logs to true resistivity as an average over an interval, if the bed is thick enough to allow the tool to read properly, but in the siliceous shales the bedding is far too thin for these tools to be valid. Other methods exist for deconvolving these curves based on tool response physics, but any of these resistivity-log corrections were beyond the scope of this project and were not done. Water saturation was determined using the relationship:


Swn = (Ct *Fo - (Swb *(Ccw Cw)) * Swn-1
_________________________
Cw

where:

Fo = Dual Water formation factor
Ct = 1/Rt = true formation conductivity
Cw = 1 /Rw = formation water conductivity
Ccw = clay bound water conductivity
Swb = fractional portion of total porosity saturated with clay bound water

This is the published form of the Dual Water equation as presented in the Schiumberger literature.

The calculation of water saturation is discussed further in Appendix IV, Section A.IV.1.3.

5.0 RESULTS OF THE STUDY

This project was not typical of most reservoir descriptions in that the results were not intended to be used for any economic purpose within the context of the project itself. The main result of DPI's portion of this study is the digital log database, the analysis, the geological correlations, and the methodology for actually performing the work. Some observations and conclusions can be drawn from the study, however, and they are discussed below.

5.1 OBSERVATIONS AND CONCLUSIONS

A model has been developed to allow the use of old E-logs and other wells lacking porosity data in an integrated reservoir description project for diatomaceous reservoirs. Such a model must be calibrated to each individual field and reservoir by including all available modern log and rock properties data. It is also important to note that the E-log only model depends heavily on a geologically-based rule set. The geologist/log analyst is performing an interpretation that must come form personal knowledge of the area when he/she sets up an old E-log for processing, so it is best if the petrophysics is done by the project geologist. If this is not possible, the petrophysicist should at least work very closely with the geologist when performing this kind of field study.

Nearly all basins in this country have a large number of old E-log wells, even if they also have many newer wells with porosity logs. The E-logs are always used by the geologist for correlation, even if they are ignored by the petrophysicist in performing log analysis. This approach can serve as a guidepost for the integration of the old log data into the quantitative reservoir description.

The Miocene diatomaceous section analyzed in this project is widespread throughout the southwestern San Joaquin Valley. The reservoir description work resulting from this effort will have application to a much larger area that is prospective for oil production. With the recent increase in oil price and level of effort in California, these results may have a significant impact on future activities in this part of the basin.

There were anomalies in the water saturation results that suggest water resistivity variations exist within the study area that could not be identified with the data available. This is probably not surprising, since the Monterey was probably deeply flushed with meteoric water during exposure at the end of Miocene time. Also, it is well known that diagenesis of feldspathic sands in the San Joaquin Valley has a significant effect on the chemistry of connate waters. Either or both of these processes could have large and unpredictable effects on water resistivity within the reservoir.

5.2 OUTPUT DATA INCLUDED WITH THE REPORT

The output product of this study delivered to MTU by DPI consists of several kinds of digital data and this written report. These deliverables are discussed below.

5.2.1 FINAL REPORT

The final report consists of a brief summary of each analysis step and issue involved in the study, followed by a full discussion included in the Appendices. Appendix V contains all figures and illustrations to which the text refers.

5.2.2 FINAL DATABASE

The final database for this project area consists of raw curves, depth shifted, and environmentally corrected curves, and a number of computed curves. The computed curves fall into two families: final model outputs and intermediate process curves used as model inputs. These are tabulated and defined in Table 3.

Other digital data delivered include a spreadsheet of well information and correlations by well. Previously DPI has delivered a set of hand-drawn cross sections and maps to MTU for use in setting up computerized displays in GeoGraphix and other systems.

6.0 DISCLAIMER

All interpretations are opinions based on inferences from electrical or other measurements made by third parties and we cannot, and do not, except in the case of willful negligence on our part, be liable or responsible for any loss, costs, damages or expenses directly or indirectly incurred or sustained by anyone resulting from any interpretations made by any of our officers, agents, or employees. These interpretations are also subject to our general terms and conditions as set out in the current Digital Petrophysics, Inc. price schedule, which are incorporated at this point as though set forth at length.




Deborah M. Olson William R. Berry II
California Registered Geologist #4198 California Registered Geologist #3903


APPENDIX I

Background Information

AI.2.1.1 ANALYSIS APPROACH

The integration of all available reservoir information, viewed from a strong geological/production engineering standpoint, combined with modern and theoretically sound petrophysical analysis techniques is the overall philosophy of field study management that guides DPI's approach to projects such as the Pioneer area study. The basic goal of the analysis is the development of a field function: a petrophysical model that can be used to derive a consistent set of computed parameters for every well. If the field function is properly calibrated and if the database is consistent, it can then be used to interpret other wells in the project area that are determined to be similar.

Every effort is made to construct a consistent database for the project, using uniform curve nomenclature, the same preprocessing procedures such as depth shifting and environmental corrections, and careful normalization to adjust tool responses. These steps are all described in detail in this report in the various sections included in Database Preparation, Section 2.2.2, and Appendix III. The key wells from this database, usually core wells, are used for calibration of the model, and parameter selection is performed and verified on them. This phase of the project is very time-consuming, but the model analysis can be quickly extended to the rest of the field wells when the database is consistent, thus saving a great deal of time on the processing of the entire project.

No field study database contains definitive information for all necessary reservoir parameters required by a petrophysical model. Most of these are derived from laboratory analyses on core material. These analyses must be graded on the basis of the procedure used, sample quality, and overall validity of the data. Most of the model parameters are based on the geological characteristics of the reservoir, such as mineralogy, grain size, sorting, and stratigraphy. DPI utilizes geologist/petrophysicists experienced in oilfield production operations for selecting model parameters. Even where excellent core data exist, the ultimate model calibration must be performed by the experience of the analyst, based not only in previous petrophysical work but in a background of sedimentary and reservoir geology. The clay and porosity transforms in particular are judged by how well they fit the known mineralogy, depositional environment, and diagenetic history of the sediments.

It is also very important whenever possible to incorporate local expertise on the geology and reservoir conditions from staff personnel who are most familiar with the field. In the case of this project, DPI staff had already worked in the area and had a good knowledge of the reservoir.

DPI has developed a set of detailed guidelines for field study management. All the necessary steps for execution of a field study are listed in their proper order. However, the system is flexible and can accommodate additions and changes in project direction and scope as technical issues are identified that alter the original plan. Without the guidelines, steps could easily be forgotten, performed incorrectly, or out of sequence when events occur that disrupt project flow.

Nearly all field studies include several elements that are almost research in nature, and often result in new ways of viewing the reservoir. These are explained fully in DPI's final report, along with a detailed record of all procedures, events, and results from the study. The product of this is an interdisciplinary tool that contains raw data, petrophysical computed parameters, and an interpretation of the analyzed data.

AI.2.2.1 PETROPHYSICAL MODEL

AI.2.2.1.1 CHOICE OF MODEL

Classic approaches to shaly sand analysis such as SARABAND's laminated-dispersed logic (Poupon et al., 1970) are unstable under a variety of conditions including very low porosity gas sands, feldspathic sand and silt sequences, and heavily cemented sands. Without logging tools that recognize thin beds, such as raw dipmeter or various micro resistivity measurements, it is not possible to distinguish between laminated and dispersed shales. Therefore, a dispersed-shale model was chosen for this study. Of these, the Dual Water and Waxman-Smits models are preferred because they are theoretically sound. Insufficient information was available to calibrate the Waxman-Smits model properly, so the Dual Water model was used in this study.

AI2.2.1.2 POROSITY MODEL

Porosity was determined from the neutron-density crossplot method for shaly sands for all the wells with porosity logs. A matrix density curve was computed and used instead of a fixed point in this crossplot calculation, since the matrix endpoint varies in diatomaceous rocks from less than 2.0 g/cc up to more than 2.75 g/cc, depending on the degree of diagenesis, clay content, and dolomitization. Hydrocarbon corrections were applied on an iterative basis to the density and neutron data for computation of total and effective porosity. The correction is computed by iteration until the porosity change is within a certain tolerance (or until a certain number of iterations have been done). The overall correction applied was that for a low-gravity API oil, which results in very little correction. In diatomaceous rocks, neutron-density crossover does not necessarily indicate gas, but is merely a function of lithology in Opal-A and Opal-Ct.

The calculation of the hydrocarbon correction was complicated by the absence of an Rxo measurement in most of the wells. Therefore, Sxo was computed from Rt using a quadratic solution of the Dual Water equation and an exponent as follows:

Sxo = Swx

AI.2.2.1.3 SATURATION MODEL

Where 'n'=2, the Dual Water model can be solved quadratically. Where 'n' <> 2, the solution is iterative. The model repeats the iteration until the solution converges within an acceptable user-defined limit, or until a set number of iterations has been performed.

The choice of electrical properties for the saturation model was quite a complex and involved procedure due to the unusual nature of the reservoir, the small amount of useful core data, and the problems to be solved. The porosity exponent "m" and the saturation exponent "n" were both variable functions of lithology. Also, the Rw value was difficult to identify, and it is likely that there are some water resistivity variations that could not be identified with the available data. Calibration procedures are discussed in detail in Appendix IV, Section AIV.2.2.4.3.



APPENDIX II

Characteristics of Diatomaceous Rocks

A.II.1.0 MONTEREY SHALE FACIES AND PRODUCTION CHARACTERISTICS

The characteristics and evolution of the siliceous shale facies of the Miocene Monterey formations are complex and have caused consternation among many operators since they were first found to be productive. The Monterey can be found productive at any depth. However the characteristics of the reservoirs vary dramatically with depth and clay content. A clear understanding of the depositional and diagenetic process involved in the formation of these reservoirs is necessary in order to appreciate the potential of the biogenic shale reservoirs at Pioneer Anticline and elsewhere in the San Joaquin Valley.

A.II.1.1 FORMATION OF A DIATOMACEOUS RESERVOIR

Diatoms are marine planktonic plants that form their shells from hydrated amorphous silica, or Opal-A. At the time of deposition, the hydrated amorphus silica is composed of one molecule of SiO2 bound with a number of H2O molecules. Under certain conditions diatoms dominate the plankton in open-sea areas and, where clastic input is minor, deposits of nearly pure diatomite can be formed admixed with a small amounts of silt and clay. The combination of normal intragranular porosity and the void space inside the hollow diatom shells can yield a bulk porosity of up to 75% in unaltered (Opal-A phase) diatomite. However, due to clastic contamination and compaction at even shallow depth, the porosity is normally in the 55%-65% range. The grain size is very small and the permeability is low, usually less than 1 millidarcy. Figure A.II-1 shows a scanning electron microscope (SEM) photograph of pristine Opal-A diatom frustules. The rock is rigid but friable. With the high matrix porosity, reservoir fluid storage potential is very large.

Under moderate heat and pressure, the unstable amorphous opal (Opal-A) of the diatom tests begins to convert to a microcrystalline form of quartz called Opal-Ct (Cristobalite). Through compaction and morphological changes related to the formation of Cristobalite, most of the diatom tests lose their intricate structure and take on the appearance of watered down corn flakes. Thus, most of the individual diatoms are no longer distinctly recognizable as fossils. There is a consequent large loss of rock volume and porosity, although a typical Opal-Ct still has a porosity of 40%-45%, high by conventional standards. Such rocks have significant matrix reservoir storage. Figure A.II-2 shows an SEM photograph of Opal-CT.

With greater heat and pressure, the Opal-Ct converts to quartz, resulting in an additional loss of rock volume and porosity. With enough heat and pressure, very clean diatomites end up as glassy chert with no matrix porosity. Under the right tectonic conditions, the glassy chert can be extensively fractured, thus providing a very high permeability reservoir. Such reservoirs may have little or no matrix storage but can have exceptionally high fracture porosities. Quartz phase rocks with admixed clastics retain some matrix porosity, and consequently have potential matrix storage for hydrocarbons. Intermediate stages of diagenesis yield a rock with moderate to low (15% to 20%) porosity which is very brittle and readily fractured. The term porcelanite is generally applied to quartz phase diatomite having low to moderate amounts of clay and moderate porosity.

Below the Opal-Ct phase window, the rock alters gradually to quartz phase, generally over a long interval. At Lost Hills and South Belridge, the quartz transition in the clean facies starts around 4,000 feet subsurface. Remnant Opal-Ct may be present in some places as deep as 5,500 feet, but the majority of the rock having low clay content is in the quartz phase by 5,000 feet. In some more tectonicaly active structures, prior depth of burial and subsequent erosion have brought Ct and quartz facies to shallower depths or to outcorp. If the diatomite was relatively pure, its quartz-phase equivalent endpoint is glassy chert. However, the transition from porous quartz phase clean diatomite to glassy chert can be over many thousands of feet, depending upon temperature and pressure. In the San Joaquin Valley, there is very little glassy chert but there are vast volumes of brittle porcelanite that was originally diatomaceous shale having low to moderate clay volume. The shale usually has streaks of pure porous quartz or chert, and often retains matrix porosity to relatively great depths. Quartz-phase rock retains both micro- and macro-fractures much better than Opal-Ct phase rock.

Recent developments around the San Joaquin Valley and better understanding of existing production from Monterey-type reservoirs indicate that the quartz-phase reservoirs can be excellent producers under certain circumstances. The rock must be clean enough to be brittle and must be subject to significant tectonic stress to propagate extensive macro fractures. Abundant additional reserves are possible where the rock retained sufficient matrix porosity and/or micro fractures to store hydrocarbons.

Production from the Monterey in coastal oil fields is often from fractured glassy chert having little or no matrix storage. However, fracturing in these reservoirs is so intense that the fracture storage is much higher than that found in most San Joaquin Valley deposits.

The transition between diagenetic phases is gradational and is partially dependent on clay content. The Opal-A/Opal-Ct transition in low clay environments usually occurs between depths of 1700 - 2500 ft. The presence of clays retards the Opal-A to Ct conversion, so in a section with interbedded clastics the transition could be stretched over an even longer interval. The Opal-Ct/quartz transition is more variable in depth and can extend as a mixed facies over several thousand feet. However, for clean diatomite (low clay), the transition occurs between 4,000 and 5,000 feet along the west side of the San Joaquin Valley.

Diatomaceous rocks become more brittle as they reach higher diagenetic grade, culminating in pure quartz cherts that shatter like glass. Brittleness and susceptibility to fracturing decrease with increasing clay content. When these diagenetically embrittled rocks are deformed in the young, tectonically active structures common on the west side of the San Joaquin Valley they fracture readily. Therefore, fracture permeability is common and greatly enhances the productive capacity of the rocks. Fracture stimulation in wells is often used in these reservoirs to facilitate the interconnection of the existing microfracture network and access the large matrix and micro-fracture storage capacity.

A.II.1.2 ANALOGOUS SILICEOUS SHALE RESERVORS IN THE SOUTHWESTERN SAN JOAQUIN BASIN

Nearly all reservoirs of upper to middle Miocene age in the Southwestern San Joaquin Valley are either turbidites or biogenic siliceous shale. They are found in formations commonly referred to as the Reef Ridge, the McClure/Antelope Shale, and the McDonald Shale. There is economic production from the siliceous shale facies of one or more of these formations in fields such as Lost Hills, North and South Belridge, and Buena Vista Hills. Some of the reservoir characteristics of these formations are summarized below.

A.II.1.2.1 REEF RIDGE FORMATION

The Reef Ridge (Miocene) consists of moderate to very clayey diatomaceous shales. In the crestal portion of the Lost Hills and South Belridge fields, the Reef Ridge is in the Opal-A phase and responds well to fracture stimulation. Further down the structure, the transition to Opal-Ct results in lower productivity in the clay rich zones. However, in some intervals having lower clay content, the Opal-Ct responds with good production rates of oil and associated gas.

A.II.1.2.2 MCCLURE/ANTELOPE SHALE FORMATION

The McClure (Antelope) Shale (Miocene) is similar in nature to the Reef Ridge. Deposition in the vicinity of the Lost Hills and South Belridge fields during Reef Ridge and McClure times was dominated by diatom blooms during periods of varying clastic influx. Diatom bloom cycles which were contemporaneous with periods of low clastic influx resulted in relatively clean diatomites. Late in the McClure times, diatom deposition was high and clay influx was low, thus yielding several hundred feet of fairly clean diatomite with low clay and silt content. At Lost Hills, this clean interval is the Cahn Zone. The occurrence of limestone or dolomite lenses is more common in the McClure than in the Reef Ridge.

A.II.1.2.2.1 CAHN ZONE OF THE MCCLURE/ANTELOPE SHALE FORMATION AT LOST HILLS

DPI classifies the Cahn Zone as being the low clay diatom rich interval which is easily correlated on neutron-density porosity logs over the entire Lost Hills structure. This zone has a very low clay content and represents a period of locally intense diatom deposition with very low clastic influx. The Lost Hills Cahn Zone has similar characteristics to the clean intervals at Pioneer. The Opal-A/Opal-CT and the Opal-CT/quartz phase boundaries in Lost Hills have subsea depth windows of 1,800 to 2,300 feet and 4,300 to 5,000 feet respectively. At Southeast Lost Hills there is strong evidence that individual well production is a function of the thickness of the clean facies and the degree of natural fracturing.

A.II.1.2.2.2 MCDONALD SHALE RESERVOIRS AT LOST HILLS

The McDonald Shale is productive on the southeast plunge of the Lost Hills anticline. The clean diatomite facies in the McDonald may be clayer than the clean facies in the Cahn zone above. Also, the McDonald contains a great number of limestone or locally dolomitic layers ranging in thickness from a few inches to a few feet. These layers may be locally fractured and contribute to production

A.II.1.3 FRACTURING

The more prolific biogenic shale reservoirs of the Monterey produce from a system of natural micro and macro fractures which drain the primary pore system. The macro fractures are tectonically induced. The micro fractures can be derived from both tectonic and diagenic causes. Micro fractures, and to a lesser extent macro fractures, can be healed or closed off by precipitation of cementing agents or by plastic deformation of the matrix.

A.II.1.3.1 MICRO FRACTURING

Micro fractures can be derived from both tectonic and diagenic proceses. The diagenic transformation of Opal-A to Opal-Ct and finally to quartz creates a material balance problem. The transition from Opal-A to Opal-Ct results in several changes to the rock. Opal-A (SiO2*nH2O) gives off most of its structural water resulting in an increase of matrix density from as low as 1.8 g/cc to a range of 2.5 to 2.65 for Opal-Ct. At the same time, matrix porosity is reduced from about 60% to about 45%. This results in a net reduction of total rock volume (inclusive of pore space). This results in a net expulsion of large amounts of water. As these biogenic facies are nearly always encased in clastic- dominated clay-rich shale sequences, complete expulsion of the water with increasing depth of burial is not always accomplished. Thus, microscopic discontinuities in the seemingly "welded" Opal-Ct are held open or "wedged" by hydraulic pressure. Thus, microfractures are created and preserved by the diagenetic phase transformation. Most Opal-Ct intervals tend to be overpressured, regardless of their structural setting. The trapped fluids have preserved the micro fractures and are the cause of the overpressure.

In cleaner Opal-Ct zones, the rock is more rigid than in the clayer intervals. Higher concentrations of clay cause the Opal-Ct facies to deform plastically when exposed to moderate tectonic stress. When exposed to severe tectonic stress, the clayey Opal-Ct will exhibit both micro and macro fracturing, but readily heals. Cleaner Opal-Ct will exhibit failure (fracturing) at lower stress. However, even fractures in the cleaner Opal-Ct facies will heal with time if hydraulic wedging is insufficient to hold the fractures open.

With the transition from Opal-Ct to quartz, there is a further decrease in rock volume due to a reduction in porosity from 45% to less than 30% and a simultaneous further increase in matrix density due to the final expulsion of any residual structural water. Again, the rock must expel water or sustain a further increas in internal pore pressure. The net effect is that much of the micro fracturing which originated with the first diagenic phase change and any subsequent stess induced micro fracturing can be further preserved by the hydraulic wedging. Again, most quartz phase biogenic shales in the subsurface are overpressured regardless of their relationship to structure.

A.II.1.3.2 MACRO FRACTURING

Most of the natural macro fracturing that is encountered in the diatomaceous shales is tectonically induced. Because most of the west side major structural features such as the Lost Hills anticline, Belridge, Pioneer, and others are young and growing structures, any structural cause of fracturing that is present can be thought of as an ongoing process. The same or similar structural mechanisms govern growth of most of the major structures paralleling the San Andreas Fault. This type of continued deformation implies that the structure is still growing and would favor the preservation of existing fractures and propagation of new ones.

A.II.1.4 PRODUCTIVE POTENTIAL OF BIOGENIC SILICEOUS SHALE RESERVOIRS

Wells drilled into a naturally fractured biogenic siliceous shale reservoir in the quartz phase should have about the same productive potential as such reservoirs as the ASZ (Antelope Shale Zone) at Buena Vista Hills. Wells drilled into biogenic siliceous shale reservoirs lacking an extensive natural macro fracture system will have about the same productive potential as the Cahn zone at southeast Lost Hills. Other fractured shale analogs also exist and are further discussed in this report. These pools, both in the San Joaquin Valley and on the coast, have produced very large quantities of hydrocarbons at very high rates.

A.II.1.4.1 FRACTURE STIMULATED PRODUCTION AT LOST HILLS

There is abundant production from diatomaceous rocks in the Lost Hills Field. In the absence of extensive natural macro fractures, the biogenic siliceous shale reservoirs may produce only a few barrels of oil and a few MCF of gas per day. However, massive (250,000 to 500,000 # frac sand) hydraulic fracture stimulation results in very high initial production rates and cumulative production ranging from 100,000 bbl to 400,000 bbls per well. Fracture stimulated production has been obtained from wells completed in both the Opal-Ct and quartz phase facies. There are many high volume producing wells in the quartz-phase Cahn Zone in the Southeast Lost Hills pool.

Opal-A reservoirs in the Etchegoin and Reef Ridge are produced through fracture stimulation by several operators along the anticlinal axis a few miles south of the structural culmination. These reservoirs have moderate to high oil saturations. Naturally fractured "Cahn chert" has been produced for many years by Chevron through slotted liner completions in the central portion of the field. In 1913 Chevron drilled a well into a naturally fractured interval of the Cahn Zone. The well produced at least into the 1950's with a gas/oil ratio of about 1500 SCF/Bbl and a water cut which increased from 50% to about 90%. Between 1949 and 1952 an additional 26 wells were put on production.

The next significant drilling program occurred between 1977 and 1982 when the number of Cahn completions jumped from 36 to 259. This drilling boom was a result of what was considered a major new pool discovery, the southeast Lost Hills field extension. The production consists of light oil and significant gas from somewhat naturally fractured siliceous shales in Opal-Ct and quartz phase. Commercial production rates were achieved by large-scale hydraulic fracture stimulation. The best wells in this pool had IP's of as much as 600 barrels of 40-degree oil per day, with gas and water. The completions usually consist of a multi-stage frac over an interval of 500 to 1,000 feet. Each frac stage is about 200 feet. The most successful frac jobs put away over 250,000 pounds of frac sand. The pool has had a history typical of a naturally fractured reservoir with good matrix storage capability: the decline rates were very high initially, but soon flattened to production of around 200 BOPD for the best wells and 50-100 BOPD for lesser wells. After the initial decline, most of these wells exhibit a decline rate of less than about 5%. The deepest wells on the southeast plunge are on 10 acre spacings. Originally the reservoir was somewhat overpressured and the IP rates declined both with the decrease in reservoir pressure. Later wells which were drilled in marginal positions to the main pool showed lower IP's as well.

Chevron and Mobil, as well as other operators, are currently active in Lost Hills, at a time when they have severely curtailed their activities in most other fields in the San Joaquin Valley. They are clearly seeing a good return from investment in light-oil producing Opal-Ct and quartz-phase diatomaceous shales in central and southeast Lost Hills.

Commercial rates of production have been obtained from wells drilled to nearly 9,000 feet in the siliceous shales of the Southeast Lost Hills pool. These wells required massive hydraulic fracture stimulation to achieve good results. Most of these wells are located on the broad and less intensely deformed southwestern flank of the southeast plunge of the structure and do not appear to have naturally existing large scale natural fractures. The greater success rate of the wells in this area is due to the thicker net clean facies in the Cahn zone and the McDonald cherts.

The most successful frac completions tend to be in the cleanest portions of the Cahn Zone and McDonald cherts. Wells with thicker intervals of clean diatomite tend to produce at higher rates and have flatter declines. Several operators are evaluating plans to drill horizontal wells in the southeast portion of the field.

A.II.1.4.2 PRODUCTION ANALOGS IN OTHER FIELDS

During the early 1950's the Antelope Shale Zone Unit at Buena Vista Hills was developed on 40 acre spacing. These wells encountered 200 feet of naturally fractured siliceous shales at depths of 5,000 to 6,000 feet. During the flush production of the ASZU unit, these wells produced 200 bbls per day oil plus significant gas. During the flush production at BVH, there were no attempts at fracture stimulation.

A.II.1.5 HIGH WATER CUTS COMMON IN BIOGENIC SILICEOUS SHALE RESERVOIRS

The combination of the diagenic processes and the high surface area to volume ratios associated with the matrix porosity in any of the three diagenetic phases results in high water cuts in most Monterey completions. Reservoirs having light oil and higher gas oil ratios may actually have gravity segregation as a function of pore geometry. Gas and lighter oils may have entered pore systems having pore throats too small to pass higher viscosity crudes. Many core descriptions refer to dark staining (suggesting heavier oil) near fracture planes but lighter oil and brighter fluoresence in the less fractured matrix. Thus, it is possible that the lighter ends have penetrated into the portion of the rock having the smaller pore throats. These micro porous systems have higher surface to volume ratios, and thus will have higher water saturation. When the reservoir pressure is lowered, the gas expands and displaces significant volumes of water from the micro pore system. It is not uncommon to have immediate 50% water cuts in intervals far from the oil water contact.


APPENDIX III

Database Preparation

A.III.1.0 DATABASE PREPARATION AND EDITING

The sequence of steps in the preparation of the database is critical to the timely and efficient analysis of the data. The following sections describe the sequence and methods employed in the preparation of the database.

A.III.1.1 DATA LOADING

All log data were digitized from paper prints for this project. They were loaded to the Petrolog program on the computer and made available for further processing. Header data off the logs were also loaded to Petrolog. Well information was entered to a spreadsheet for use by the geologists and log analysts on the project.

A.III.1.2 DEPTH SHIFTING

All depth shifting was performed interactively. Only the wells with core and porostiy logs required depth shifting. In all cases, the shallow resistivity curve was taken to be tool zero. Porosity logs were shifted using standard methods. Core data were shifted by comparing the linear-interpolated core porosities to a preliminary computed log-derived porosity. After depth shifting, the core porosity can better be used to calibrate the analysis.

The question is sometimes asked why the core data are linear interpolated before depth shifting. When depth shifting, the analyst uses visual comparison to determine a target depth for a given part of an input curve, so a depth pair is generated. The depth shift algorithms used in most analysis programs are of the "rubber band" type; they use successive depth pairs to compute a positive or negative shift for each depth point and then move it to the nearest database increment (commonly one/half foot). Point data, such as input core data, are usually moved by these methods into the spaces between depth increments and are lost. A linear-interpolated set of core data solves this problem since data are not lost from a continuous curve.

A.III.1.3 ENVIRONMENTAL CORRECTIONS

Environmental corrections must be applied prior to the normalization step and must be considered for each curve. The chart-book corrections appropriate to the service company were made as required. Gamma ray, density, neutron, and induction curves were all corrected according to the chartbook algorithms included in the Petrolog program. The SP curves were baseline shifted to produce a curve called SBL.

A.III.1.4 NORMALIZATION

Normalization becomes an issue in petrophysical studies whenever more than one well is to be evaluated. The goal of a normalization procedure is to calibrate all the log responses to a single standard. An absolute standard such as a dolomite or massive salt is preferred, but in the absence of this a relative standard is chosen that is representative of what a majority of the logs already read; the database is therefore calibrated to itself. Variations in tool readings are the result of calibration error, differences between service companies, variations in tool design or (especially in the last ten years) differences in software processing applied to the raw tool data. Real spatial variations also exist, especially in complex geological environments, and the normalization process must take these into account and avoid editing them out of the dataset.

The dataset for this project was only partially normalized in terms of the procedures described below because the full-suite wells were scattered over several sections and the differences in diagenetic grade in the siliceous shales masked the effect of miscalibrated logs. In particular, the gamma ray curves showed such variation that it was unclear how to normalize them. They were not used in computation of clay volume. However, the theory and procedures of normalization are discussed here since this project will also be used as a guide to work elsewhere.

A.III.1.4.1 NORMALIZATION DEFINED

The term 'normalization' as found in the literature has several meanings in log analysis. In this report, 'normalization' is used in the context of a field study where a statistically valid number of wells in the study area are compared to identify those variations in log response that can be attributed to log calibration errors. The logs are corrected, or normalized, by applying the appropriate shift to each curve as calculated from the normalization process. If multiple logging runs are present, it is necessary to normalize each run, but a single normalization applies to an entire logging run regardless of how many zones are included. Analysis parameters that are determined from calibration to known data points, such as coreholes, can then be applied with confidence to the entire project. Therefore, areal variations in reservoir properties that are seen in the results can be attributed to real-world differences and not to spurious log errors.

To perform a valid normalization, the log data must first be verified, edited, depth shifted, and environmentally corrected. These tasks have been described elsewhere in this report. Geological correlation should also be done prior to normalization, so that the same stratigraphic interval can be selected in each well for comparison.

As in the Pioneer area, most field studies in sand/shale sequences have no lithologies with a known and consistent log response (such as carbonates or salt), so all normalizations must be performed as relative shifts to a locally defined standard. A local standard is usually established by examining all the wells in the study and identifying the log response of a majority of the wells sampled. It is important to use a stratigraphic interval that includes both sands and shales so that both ends of the log response spectrum are represented.

The standard is not an absolute value, however, since log response is also influenced by geological variation and changes in fluid saturations within the reservoir and over time. A mapping approach to the normalization problem for the porosity logs is recommended and described in this report. DPI has performed statistical research as part of previous projects (not a part of the current project) to test the usefulness of various statistical techniques in evaluating the logs for normalization. DPI has also tested the effects of using various stratigraphic intervals and examining only selected lithologies within an interval. The recommended normalization procedures reflect the results of this research and also of the many field studies DPI has performed throughout the world. These procedures are discussed below under the sections for specific curves.

A.III.1.4.2 GR METHODOLOGY

Histograms of GR values were made over the selected stratigraphic iinterval and compared. A group of wells were chosen to be the standards whose histograms were very similar and the other wells were compared to these. For most wells logged by the same company, the range of gamma ray response was similar and the only normalization required was a simple shift of several units. Wells logged by different companies usually require larger shifts, and sometimes a range expansion. This does not mean that the logs are bad or wrong, only that the tool design is different. Normalization takes care of this difference in response betwen companies so that the same analysis paramters can be used throughout the study. Wells that require a range expansion can usually be normalized by applying an equation of the form y=mx+b (straight line).

The gamma ray curve responds to changes in the amount of radioactive elements in the formation and borehole. The type of clays in the mud system and their inherent radioactivity varies from well to well. Variations in mud radioactivity will shift the histogram to the left or right when compared to other wells. Variations in the formation lithology will cause relative changes in the shapes of the histograms, with shaly intervals having more high gamma ray and sandy intervals having more low gamma ray. In most cases, it is not necessary to account for possible geological variations that could affect the gamma ray response and the logs can be shifted to a median value. The normalized gamma ray can then used to compute a clay volume to investigate the effects of clay on the responses of the porosity and induction curves.


A.III.1.4.3 POROSITY NORMALIZATION

Ideally, an interval barren of hydrocarbons should be used for porosity log normalization, since gas and light oil can severely affect the response of the density and neutron curves. If this is not possible, then the potential effects of hydrocarbons must be kept in mind when comparing logs from different parts of the structure and from different dates. The date is important since hydrocarbon composition and saturation change during reservoir depletion due to pressure changes, waterflooding, gas cap expansion, and other causes.

In many field studies, there is a dominant logging company. If this is the case, the standards should be chosen from this company's logs only, and one should expect to shift the logs from other companies to these standards. As for the gamma rays, differences between logging companies do not mean that one company is better than another, only that the tool calibration standards are not the same. The normalization and parameter selection procedure can and do compensate for any differences between logging company.

A.III.1.4.3.1 DENSITY METHODOLOGY

Histograms of the density curve were made over the selected interval and compared. Initially, the correct log response was unknown. Most wells had similar widths to their ranges, but the distributions were shifted laterally from one another. Some of these shifts were likely calibration errors, but most were variations in porosity. Therefore, the logs cannot simply be shifted laterally to match the standard since true geologic variations will therefore be shifted out. To address this problem, the data were mapped.

The simplest kind of map, one of the means of the distributions, is somewhat misleading, since the mean will be different in two wells that have the same log response if one has more sand than the other (Fig. A.III-1). A better method is for the analyst to identify the most common density curve type in the dataset, then compare the histograms to the type by hand and determine a shift from the standard, which is then mapped. This is done by overlaying all the histograms and finding a group whose members match. Histograms are judged to match if the leading and trailing edges have the same numeric values (the range is the same) and if the overall shape is similar. Changes in sand/shale ratio can affect the relative size of modes at the ends of the histogram, which can radically shift the mean value, but an experienced analyst can readily recognize a matching histogram from a shifted one.

Typically, a group of wells are identified whose histograms matched. Histograms for other wells in the project are shifted to either side of them. These wells are used as the standards to which others were compared. A +/- shift was then determined for each well as compared to the standard. These shifts were then mapped to look for non-systematic variations over the area of interest.

If the reservoir had an initial gas cap, one would expect to see light density and low neutron values on top of the structure, grading down the flanks to heavier density and higher neutron. If reservoir fluids had changed through time, one would see trends based on drilling dates. There may be other trends related to geological variation and are supported by several wells. Such trends as these are not tool error and should be left alone. Instead, wells that stand out as anomalous within these trends should be inspected and shifted to fit the trend. Note that an effort to shift all the density logs to a common standard would erroneously eliminate these real porosity changes.

A.III.1.4.3.2 NEUTRON METHODOLOGY

As for the density and gamma ray, histograms of the neutron curve from each well were made over the selected interval and compared. Even more than the density curve, the neutron is heavily affected by hydrocarbons and by changes in sand/shale ratios, so a simple comparison of histograms and a linear shift to the standard will be invalid since real lateral changes in the reservoir will be eliminated. Again, as for the density, the neutron data must be mapped.

The simple map on the mean of the neutron histograms shows more information about geology than it does tool response. In fact, such a map typically shows a combination of areal geological changes and increasing gas effects (lower neutron) in later wells.

A.III.1.4.4 INDUCTION METHODOLOGY

In field studies in clastic reservoirs, normalization of the induction logs is frequently ignored because there are seldom lithologies in clastic sequences with known absolute resistivity values (such as anhydrite, salt, or non-porous carbonates). However, some of the anomalous saturation data obtained from analysis of certain field wells may be due to induction log calibration problems. Therefore, DPI recommends that the deep induction curves be included for normalization. Since the curves are originally recorded in conductivity, all induction curves were converted back to conductivity for normalization.

For perspective, the standard tool error for an induction log is +/- 5 mmhos. At 10 ohmm, this equals +/- 0.5 ohm, which in a sand with 18% - 20% porosity can equal approximately +/- 1.5% in saturation units. If the log has an error of as much as 25 mmhos, the equivalent saturation error will be approximately +/- 5% in saturation units. At higher resistivities, the effects of these errors is magnified. For example, in a 100 ohm sand, a log with a +25 mmho error would only read 28.6 ohms.

The deep induction curves within the hydrocarbon-bearing zone are far too much affected by the oil to perform any meaningful normalization, so it is especially important to select a barren interval. Ideally, a shale section should be chosen, since the errors in conductivity range will be most obvious in low resistivity intervals and any hydrocarbons present will have minimal effect.

APPENDIX IV

Parameter Selection


A.IV.1.0 PARAMETER SELECTION

Discussed in this section are the detailed procedures involved in parameter selection for the Pioneer Anticline project.

A.IV.1.1 CALCULATION OF VCLAY

Gamma ray, SP, and neutron-density crossplot are usually the best available clay indicators in a typical shaly-sand sequence. However, all of these measurements include contributions from both clays and from other reservoir conditions whose effects are difficult or impossible to quantify. Also, a large number of the wells in this project had only E-logs and therefore only SP and resistivity curves available.

The gamma ray curves showed a huge variation in range both vertically within wells and between wells, and this is attributed more to changes in radioactive mineral content than to lithology changes. The gamma ray could be used for clay volume calculation in most wells, but many zones would have to be established by the analyst based on arbitrary decisions about changes in the gamma ray response. Since these curves were variable and were not available on the E-log only wells, the clay volume in all wells was computed with the baseline-shifted SP curve.

The SP deflection is a function of clay content, clay type, and the difference in ionic potential between the mud and the formation fluids. The presence of hydrocarbons can suppress the SP deflection. Rw is suspected to vary unpredictably in this field. Apparent variations in clay content based on the SP using traditional analysis techniques are therefore partially due to changes in other factors. In this study, the effects of these other factors is minimized by the analytical approach.

First, the analyst carefully baseline-shifted the SP using geological knowledge and the resistivity curve to aid in interpreting lithologies. Then, the curve responses of all SP's in the study were normalized by range expansion or contraction to a common deflection magnitude. However, the variations in SP deflection from Rw changes and other non-clay effects are still present in the curve. DPI has introduced the concept of the clean SP baseline, analogous to the shale baseline. The analyst carefully creates a curve that joins SP peaks he/she believes to be clean, and the values of this curve are used instead of a constant in the non-linear SP equation. To some extent the analyst is interpreting when setting the SP baseline, and it must be emphasized that the SP-clean curve is also an interpretation: the geologist/analyst is making a quantitative interpretation of clay volume based on his/her best assessment of the lithology of the rocks as shown by the logs and by local knowledge.

In some wells, the SP was a bad curve, and the resultant clay volume was not valid. For these wells, clay volume was computed from the shallow-reading resistivity curve.

A non-linear transform was used to compute the clay volume from both the SP and resistivity wherever they were used. This kind of transform takes into account the fact that the diagenetic clays in the sands are of a different type and character from the detrital clays in the adjacent shales. The generic transform takes the shape shown in Fig. A.IV-1; the specific shape is dependant on the parameters selected for local conditions.

A.IV.1.2 CALCULATION OF POROSITY

Standard crossplot techniques were used to compute porosity from the density-neutron logs for all full-suite wells. However, a curve was used for the matrix value instead of a constant to account for the radical changes in rock density that occur in diatomaceous sequences.

Whole core porosity data from the Cymric wells and from the Tenneco 62X-30 well within the study area were used to calibrate the porosity model. Assuming that the porosities from Dean-Stark analysis represent total porosity, the data were used to set up this part of the crossplot porosity model.

For the E-log only wells, an empirical model of porosity was developed based on these full-suite wells. It was noted that the porosity in the siliceous shales decreased with depth in a predictable way, as would be expected from what is known about phase changes and compaction in these rocks. Therefore, an empirical function relating porosity to depth was developed and applied to the E-log only wells. Since a clay volume had been computed for each one, it was possible to derive an effective porosity from the computed total porosity.

A.IV.1.3 CALCULATION OF WATER SATURATION

No information was available from the field for electrical properties (Archie 'm' and 'n'). The actual values used were determined by sensitivity analysis. Both parameters varied as a function of lithology, and were lower in clean rocks than in shales. However, the apparent changes in water resistivity throughout the study area could not be quantified and caused the water saturation calculations to be doubtful in many wells. For the E-log only wells, saturation could not be computed at all due to the lack of a suitable resistivity curve.

Little water resistivity data are available for the study area. Chemical analyses indicate somewhat variable but moderate resistivities. The Rw value used in the study was consistent with a water of about 25,000 ppm, typical for the upper Miocene in California, but in many wells the resultant water saturation is very low. Without further information about the reservoir fluids, the analysts hesitated to increase Rw until a saturation value that was deemed more reasonable was computed, since DPI does not recommend that the analyst make arbitrary decisions about parameters without field data or sound geological reasoning to support them. Therefore, the saturation data as calculated are left in the database for a later researcher to improve upon.


TABLE 2

SUMMARY OF HYDROCARBON SHOWS

T11 R23

WELL SECTION CORE SWS M.L. TEST

JERGINS 5 4 X

STEVENSON 1 6 6427-33

x7452-68 X 8516-641

MC FARLAND 1 7 X

RAS 523 10 X X PROD

353X 11 X X

LIEBHOLTZ 45-14 14 X6521-40

6873-91

7024-39 X6635

W. MINERALS 10 16 X X

HAVENSRITE 1 17 X

HAVENSRITE 7 17 NO HISTORY

MARICOPA 1 18 NO HISTORY

G.P. 54X-20 20 X

G.P. 52X-20 20 X X

G.P. 41X-20 20 X

VISALIA MARICOPA 1 20 X

WOODWARP 4 21 x 1952-62

SANTIAGO CREEK C7 21 ?

WOODWARD 35 22 X

ARCO USL 1 22 NO HISTORY

W. MINERALS 4 22 ?

W. MINERALS 6 22 ?

W. MINERALS 8 22 BLEW OUT @ 1737' OIL & GAS

RIPLEY 1 25 X 541-44

RIPLEY TRUST 3 25 X491-3003

RIPLEY 41 25 X 2840-50

RIPLEY 44 25 ?

SANTIAGO 1 25 ?

W. MINERALS 37 25 X

W. MINERALS 8 25 X265-1237

1 MW 1 23 X 1140-1075

1 MW 2 23 X

SHELL USL 1 23 X828-1082

WEST MINERALS 12 26 X

WEST MINERALS 11 35 X243-318

SW 31-12 35 X 2327

1MW-27 23 X 1346-1446